New version of ADRV9361-Z7035 SOM works on BOB, not on embedded product.

Our customer has been using some AD9361-Z7035 SOMs (the ones in the package at the bottom of this picture) & has developed a motherboard into which they plug into & all is good.  They've developed their application & FPGA firmware.

Recently they got new SOMs (the ones in the top of this picture - I know to be Rev F).  They purchased a small number of them & they all exhibit the same fault:

  1. They are working fine when plugged to Analogue Device BOB
  2. They are not powering up when connected to their motherboard.

This raises a few questions.  *Please can you confirm the Rev of the SOMs in the lower part of the picture & is there might be some difference in power requirements of the newer revision of SOM?* 

  • the customer confirms that the SOM that arrived in the box with ref C01205F is a rev E.

    The difference is between rev E and rev F2.

    They have found a difference in the BOM between the revision E and rev F2 has a resistor missing that is R49 that is the connection between the super-sequencer U7 and the 1V regulator.

    Despite the fact that two revisions have the same schematic, *power-up sequence is different due to the BOM change!* U7 is actually programmed so only ADI know the difference, the chance that just adding R49 would make the two boards behaving the same are really minimal.

     The customer asks if they simply to live with that and adapt our motherboard to the new SOM revision. 

  • 0
    •  Analog Employees 
    on Oct 29, 2020 11:11 AM 1 month ago in reply to Knb@Rfpd

    Moved to FPGA reference design community

  • 0
    •  Analog Employees 
    on Nov 23, 2020 4:53 PM 8 days ago in reply to Knb@Rfpd

    Hi,

    Sorry for the late reply. Rev F2 and rev E should be functionally identical. Changes were only made on the layout in order to improve the RF performance, and some passives which became obsolete - but were replaced with identical specs.  

    R49 should not affect the functionality because Voltage Margining is not implemented in the sequencer. The sequencer SW has not changed(source is available here)

    We ran extensive tests with both our carriers(BOB and FMC) and have not seen any issues.

    If you can share the design of the custom carrier we can review/compare in order to try to debug the issue. 

    Thanks,

    Mihai